To this day, nobody is quite sure of the exact formula for the Dark Ages’ most terrifying ‘weapon of mass destruction.’

The rise of Islam at the be- ginning of the 7th century presaged centuries of conflict with the Byzantine Empire. The jihad, or holy war, led by the military commanders of the new faith toppled one outlying city and province after another. In 634 Caliph Omar attacked Syria. Emperor Heraclius’ Byzantine army suffered defeat at Ajnadain in the same year. Damascus fell in 635, followed in rapid succession by Jerusalem and Antioch. The last remnants of Syria capitulated in 636.

Heraclius’ woes outlived him, and his successor, Constans II, inherited them. The Muslim advance continued across Asia Minor and into Africa. Cyrenaica, Tripoli, Armenia, Cyprus and Rhodes had all fallen to Islam by 654. Encouraged by their unbroken string of victories in the name of Allah, the Arab caliphs turned their attention to the very heart of the Byzantine Empire, Constantinople.

The Byzantine historian Theophanes recorded in his Chronographia that in the fourth year of Constantine IV’s reign: “[The] deniers of Christ equipped a great fleet, and after they had sailed past Cilicia, Mohammed, son of Abdelas, wintered at Smyrna, while Kaisos wintered at Cilicia and Lycia….The emir Khalid was also sent to assist them insomuch as he was a competent and bold warrior. The aforesaid Constantine, on being informed of so great an expedition of God’s enemies against Constantinople, built large biremes bearing cauldrons of fire and dromones equipped with siphons and ordered them to be stationed at the Proclianesian harbor of [Constantinople].”

Constantine IV is not remembered as one of the great military strategists of the ancient world. Nor was the empire itself renowned for the sort of martial prowess that its antecedent, Rome, had earned in the preceding centuries; it depended more upon diplomacy than weaponry to keep potential enemies from its gates. So when the Arab commander Yazid approached the imperial capital in 671, he must have been supremely confident of the outcome.

Theophanes’ chronicle for the following year’s campaign provides few details, yet it is clear that the Arabs met unexpectedly stiff resistance: “In this year the aforesaid fleet of God’s enemies set sail and came to anchor in the region of Thrace, between the western point of the Hebdomon, that is the Magnaura, as it is called, and the eastern promontory, named Kyklobion. Every day there was a military engagement from morning until evening, between the outworks of the Golden Gate and the Kyklobion, with thrust and counter-thrust.”

Two factors worked in the Byzantines’ favor. First, while the “Greek” armies were not noted for their offensive prowess, they had become masterful at holding well prepared defenses. Constantinople was a formidable citadel. In addition, it was the adopted home of a refugee from the overrun city of Heliopolis in Syria, an architect named Kallinikos. It was not Kallinikos’ skill as an architect, though, that made him valuable to Constantine, but rather the extraordinary volatile properties of the formula he is credited with inventing for a liquid that the Byzantines themselves called “artificial fire” and the Western Crusaders of a later century dubbed more famously “Greek fire.”

Descriptions of Greek fire in the historical records of the Byzantines—and of their enemies against  whom it was used—are tantalizingly brief and imprecise. A few of its characteristics, however, are more or less universally agreed upon. Incendiary weapons of various sorts had been used in warfare before, but none with the frightening effectiveness of Greek fire. It proved especially useful for the defenders of the port city of Constantinople, whose most worrisome threat was the Arab fleet, since the liquid burned even on water. The only things that would extinguish it, according to the Byzantines, were vinegar and urine. What’s more, the flammable liquid would cling to any surface, so that it would indeed, as 13th-century philosopher Albertus Magnus wrote, burn up everything it met.

According to Byzantine historians, it burned up the Arab fleet, or at least a good portion of it, compelling the survivors to flee. When the would-be conquerors returned in 718, fire literally rained down on them once again, with the same results. Not surprisingly, the Byzantines took pains to make sure that the secret of their artificial fire stayed secret. A letter written by Emperor Constantine VII to his son in the middle of the 10th century reveals the awe in which the Byzantines themselves held Greek fire, and their obsession with having a monopoly on the weapon: “[Greek fire] was revealed and taught by God through an angel to the great and holy Constantine, the first Christian emperor, and concerning this…he received great charges from the same angel, as we are assured by the faithful witness of our fathers and grandfathers, that it should be manufactured among the Christians only and in the city ruled by them, and nowhere else at all, nor should it be sent or taught to any other nation whatsoever.”

Whether those in the know really believed in their supreme weapon’s divine origins is doubtful. More likely, the story was circulated in order to enhance the mystique surrounding it, and to reinforce the fear of divine wrath against any Byzantine commander unpatriotic enough to consider enhancing his personal fortune by selling the secret. But lest anyone doubt God’s readiness to punish such traitors, Constantine VII reminded his heir: “He who should dare give of this fire to another nation should neither be called a Christian, nor be held worthy of any rank or office; and if he should be the holder of any such, he should be expelled therefrom and be anathematized and made an example for ever and ever, whether he were an emperor, or patriarch, or any other man whatever, either ruler or subject, who should seek to transgress this commandment. And [Constantine] adjured all who had the zeal and fear of God to be prompt and to make away with him who attempted to do this, as a common enemy and a transgressor of this great commandment, and to dismiss him to a death most hateful and cruel.”

As a result of the secrecy imposed upon the manufacture and use of Greek fire, historians today are nearly as mystified by the exact nature of the weapon as Yazid’s mariners were terrified by it. The weapon embodied two secrets: the composition of the combustible mixture, and the means of projecting the stream of fire toward the enemy. A number of accounts, each dating from several centuries after Yazid’s defeat, provide some tantalizing but generally unreliable testimony as to how the liquid was made. Anna Comnena, daughter of the 12th-century Emperor Alexius I, wrote in her biography of her father: “This fire they made by the following arts. From the pine and certain such evergreen trees inflammable resin is collected. This is rubbed with sulphur and put into tubes of reed, and is blown by men using it with violent and continuous breath.”

The description is clearly incomplete at best, and probably reflects the sad reality that by Anna’s day the secret had already been long since lost and even the Byzantines were reduced to guessing. The best that modern researchers have been able to confirm is that “the formula was surely based on naphtha, perhaps with resin added as a thickener. Quicklime may have been added to make it burn in water, though not to ignite it, and saltpeter may have been added to produce an explosive effect, but neither of these components was necessary.”

Greek fire was not the first incendiary weapon by any means, and the Arabs themselves had similar flammable agents in their own arsenals. But something set Greek fire apart and earned it legendary status. Exactly what this distinctive element was remains part of the mystery. A clue may be found in the fact that the ancient accounts nearly always speak of it as a naval weapon. The only reference to its being employed against land forces appears to be the result of confusion on the part of the chronicler, who seems to have mixed up the accounts of two different battles.

The use of Greek fire exclusively at sea suggests that its uniqueness consisted in its ability to burn on water and the amazing difficulty of extinguishing it. When other incendiary weapons would have to score a direct hit to have any effect, Greek fire could probably set the surface of the sea ablaze over a wide area. Even if it did not succeed in destroying wooden ships outright, a few “rounds” of Greek fire would likely have been sufficient to close narrow straits like the Bosporus for hours on end. Interestingly, Theophanes’ account of Yazid’s defeat does not explicitly credit Greek fire with destroying any Arab ships. It says only that the Arabs withdrew after an unsuccessful siege in which the Byzantines had employed Greek fire, and that the Arab fleet was “sunk by God” on the return voyage. Theophanes goes on to say: “And in the spring they set out and, in similar fashion, made war on sea against the Christians. After doing the same for seven years and being put to shame with the help of God and His Mother; having, further more, lost a multitude of warriors and had a great many wounded, they turned back with much sorrow. And as this fleet (which was to be sunk by God) put out to sea, it was overtaken by a wintry storm and the squalls of a hurricane in the area of Syllaion. It was dashed to pieces and perished entirely.”

The traditional view is that Greek fire obliterated many of the Arab vessels, but it could be that the secret weapon simply kept the enemy ships at bay until dwindling supplies and sickness and fatigue among the Arab crews made their retreat necessary. The “multitude” of casualties Theophanes wrote of might just as well have been inflicted by archers, upon whom the Byzantine military heavily relied.

In spite of all the Byzantine precautions, their enemies did finally get their hands on Greek fire. In the same letter written by Constantine VII to his son, in which he described the punishment merited by anyone rash enough to divulge the secret, the emperor admitted, “It happened once, as wickedness will still find room, that one of our military governors, who had been most heavily bribed by certain infidels, handed over some of this fire to them….” As his account ends with the unfortunate traitor being consumed by fire from heaven, it may be nothing more than another fabrication intended to strike fear into the heart of anyone who might have been tempted to accept a bribe. Even so, in 814 the Bulgars did manage to capture a large stock of Greek fire but were unable to figure out how to use it. Apparently, the incendiary liquid was not a simple weapon; employing it effectively required a good deal of technical skill. That technical knowledge comprised the second part of the Byzantine secret.

A manuscript kept in the Vatican contains an image that documents the apparatus used to deliver the fire against an enemy, at least in very general terms, far more concretely than any record of the formula itself. The image shows a small boat crewed by three oarsmen. In the bow, two additional men operate a device that seems to consist of a large flared tube from which a cloud of fire issues forth to envelop an enemy vessel.

That depiction is reinforced by a written description attributed to either Emperor Leo III or Leo VI: “The front part of the ship had a bronze tube so arranged that the prepared fire could be projected forward to left or right and also made to fall from above. This tube was mounted on a false floor above the deck on which the specialist troops were accommodated and so raised above the attacking forces assembled in the prow. The fire was thrown either on the enemy’s ships or in the faces of the attacking troops.”

The statement that Greek fire was “thrown” at the enemy has led to the assumption that a variant form of the weapon, similar to a hand grenade or what is today called a Molotov cocktail, was also used. But in context the Byzantines seem to have intended “throwing” the fire to be understood in the same sense as that of a modern-day flamethrower.

The purpose of these spigots, or “siphons” as the Byzantines themselves called them, is readily apparent. Not so obvious is the method employed to pressurize the liquid in the tube. Surely, this, too, comprised a vital element of the secret and explains why the Bulgars failed to make use of their captured stock of the substance. Suppositions as to how it was done postulate that the volatile concoction was stored in tanks below deck, where it was heated prior to discharge. If so, it must have been a delicate operation requiring quite a bit of nerve. Such a volatile concoction would not take kindly to being heated in a sealed container. Another chronicler states that compressed air was used to force the fire from the mouth of the siphons.

Whatever the exact method, it seems certain that a specialized vessel, or at least a substantially modified warship, was needed to project the fire onto an enemy ship. The equipment involved in creating the necessary pressure is probably another reason that Greek fire was used primarily, if not exclusively, as a naval weapon. The hold of a warship might be fit for storing and pressurizing large quantities of liquid fire, but it is harder to imagine how the technology could have been incorporated into an effective and portable platform for land combat. (One Byzantine source, on the other hand, does recommend its use by the garrisons of fixed defenses to protect themselves against siege towers.)

Both the formula for the liquid means of delivering it, then, seem to have been closely and the guarded technological innovations. But the most mysterious aspect of this ancient weapon has to be why the Byzantines didn’t make greater use of it. Throughout the centuries following the rout of the Arab fleet, the defenders of a Byzantine Empire that was in decline and beset by frequent attacks seem to have employed Greek fire on only a handful of occasions. As technical historian Alex Roland has noted, “During the years when Byzantium had a supposed monopoly on the ultimate naval weapon system of the medieval world, it suffered repeated and often disastrous naval reverses all over the Mediterranean Sea.” As an example, he points out that during four successive attacks against Constantinople in the late 9th and early 10th centuries, the Russians encountered Greek fire only once.

Several likely reasons suggest themselves for this very limited use. The first stems from the Byzantines’ obvious obsession with secrecy. The more the Byzantines used their super weapon, the more they risked a fire ship being captured and reverse engineered by an enemy. “Greeks” were probably more willing to lose battles and even fleets than their precious secret.

Practical aspects are also likely to have limited the weapon’s usefulness. Though no contemporary account explicitly confirms it, Greek fire seems to have been much more effective in defense than on the attack. As mentioned before, the Byzantine fire ships used defensively could probably lay down a wall of flames that would stop an oncoming fleet dead in its tracks, but on the attack, they would have had to mix with the enemy fleet, leaving themselves vulnerable to counterattacks from the rear. In such a case, a clever enemy commander would have made sure to keep the wind at his back so that the fire would tend to blow back into the attackers’ faces. In fact, the technology of the day cannot have provided any sure safeguards against accidents, and Greek fire may have been as frightening to the Byzantine sailors as to their opponents. While there is no mention in the records of a Byzantine ship immolating itself, this end seems inevitable given a frequent enough use of the weapon.

To those reasons for the limited deployment of Greek fire must be added a third and truly remarkable certainty. At some point between the 8th and 13th centuries, the Byzantines themselves lost the secret. The explanation for this extraordinary blunder—akin to the United States forgetting how to build an atomic bomb after World War II—lies in the same obsession with secrecy that kept the weapon out of the hands of Byzantium’s enemies. According to tradition, only two families in Byzantium were privy to the full knowledge of how to make and use Greek fire—the emperor’s and that of Kallinikos’ descendants, the Lamptoses. There is even a legend that an underground tunnel connected the royal palace with Constantinople’s main arsenal at Manganes, where the mixture would most likely have been made. Certainly a few chemists, shipwrights and engineers, at least, would be required to know enough to create the liquid, fire ships and siphons. But restricting the knowledge to each one’s own narrow, specialized component of the entire system would have ensured that only two individuals in each generation would know the full process involved in creating and deploying Greek fire. The chemists would have been capable of delivering the liquid to the engineers, but would have had no idea how to heat and pressurize it without blowing themselves into oblivion. The engineers would have had no idea how to build the naval platform necessary to store and deliver the liquid fire at sea. And commanders would have no authority to make or employ Greek fire at their own discretion, but could only use what the emperor supplied them. Judging from the number of military coups throughout Byzantine history, this seems to have been a wise precaution.

The dangers of this strict secrecy are as obvious as the advantages. In a very stable and peaceful society, it might have worked well, but a stable and peaceful society would have had no need to keep such a secret. In fact, Byzantium was neither stable nor peaceful, and while the tight security surrounding the weapon ensured that it would not be revealed to the empire’s enemies, it also ensured that it would sooner or later be lost to the Byzantines themselves.

In addition to foreign invaders, the Byzantine emperors had plenty to worry about at home, and relatively few of them lived to a ripe old age. The traditional reference to the secret’s being known to “the emperor’s family” does not mean a single, unbroken line of descent from father to son. The royal family changed frequently, as disputed successions and coups deposed one line after another. Between 685 and 717, a dizzying series of coups deposed a succession of rulers. First General Leontius of Hellas took the throne. He was supplanted by a fleet commander named Tiberius, who was in turn ousted by the Armenian General Philippikos. Within two years, another coup deposed Philippikos in favor of a civilian, Artemius, who was then chased out by troops loyal to Theodosius III. Theodosius held onto power for less than two years before being supplanted by the Anatolian Leo III, who finally brought some degree of stability to Byzantine politics.

Roland postulates that during one of those coups, the usurpers too hastily put the sole guardians of the secret to death, and thus sent the knowledge of Greek fire to the grave with them. It is indeed hard to imagine that anyone close enough to the emperor to be trusted with the greatest state secret of all would escape the inevitable purges when a rival took the throne by force.

It is intriguing to consider what might have been the fate of the Byzantine Empire had the secret of Greek fire  not been lost. Effective as the weapon seems to have been under ideal circumstances, however, it would be hard to imagine its having any decisive effect on history. As a defensive weapon system, it would not have helped Byzantium to recover the Western empire that had collapsed following the sack of Rome. As a naval weapon, it would not have helped the empire to hold those lands that they did reclaim.

As for its own defense, Byzantium held out against the Arab onslaught for several centuries, even without the benefit of true Greek fire. Ultimately and ironically, Constantinople’s end came first at the hands of its erstwhile Western allies, the Crusaders. In 1204 knights of the Fourth Crusade, hoping to establish a kingdom of their own, achieved what centuries of Islamic campaigns had not—the destruction of the city. Against this brand of treachery, Greek fire would have been of no use.

In 1261 remnants of the Byzantine Empire under Michael VIII recaptured the city, but the rapidly sinking fortunes of the Byzantines were irreversible. A new enemy, the Ottoman Turks, drove the final stake into the empire’s heart in 1453. It was here, at the very end, that Greek fire might have served, if nothing else, to delay the inevitable defeat for a short while.

Although it is impossible to say exactly when the secret was lost, it was surely long gone by this time. The Byzantines did employ an incendiary weapon of similar nature against the Turks, which is often referred to as Greek fire. Most likely, however, this weapon was a poor substitute, the result of a gallant but futile effort to reinvent the decisive weapon of the 7th century. Or perhaps the Turks, knowing full well the stories of the legendary ancient fire, simply assumed that any incendiary weapon the Byzantines used must be Greek fire.

Was Greek fire ever really as potent a weapon as legends and sketchy historical accounts depict? Or was its effectiveness due mostly to the terror it provoked in Byzantium’s enemies? Barring the unexpected discovery of its recipe in some still hidden archive, only the ghosts of Constantinople can say.


Bruce Heydt, a former editor of British Heritage Magazine, is a freelance author who publishes on a variety of historical topics and eras. For further reading, try Greek Fire, Poison Arrows and Scorpion Bombs: Biological and Chemical Warfare in the Ancient World, by Adrienne Mayor.

Originally published in the April 2006 issue of Military History. To subscribe, click here